change synthesizer factory api to take in aiohttp session (#318)

* change synthesizer factory api * update synthesizer constructors
vocodedev · Jul 30, 2023 · 8fb11e5 · 8fb11e5
1 parent 4ce40fc
commit 8fb11e5
Show file tree

Hide file tree

Showing 10 changed files with 47 additions and 18 deletions.
diff --git a/vocode/streaming/agent/restful_user_implemented_agent.py b/vocode/streaming/agent/restful_user_implemented_agent.py
@@ -33,7 +33,7 @@ async def respond(
     ) -> Tuple[Optional[str], bool]:
         config = self.agent_config.respond
         try:
-            # todo: cache session
+            # TODO: cache session
             async with aiohttp.ClientSession() as session:
                 payload = RESTfulAgentInput(
                     human_input=human_input, conversation_id=conversation_id

diff --git a/vocode/streaming/synthesizer/azure_synthesizer.py b/vocode/streaming/synthesizer/azure_synthesizer.py
@@ -5,6 +5,7 @@
 import re
 from typing import Any, List, Optional, Tuple
 from xml.etree import ElementTree
+import aiohttp
 from vocode import getenv
 from opentelemetry.context.context import Context
 
@@ -62,8 +63,9 @@ def __init__(
         logger: Optional[logging.Logger] = None,
         azure_speech_key: Optional[str] = None,
         azure_speech_region: Optional[str] = None,
+        aiohttp_session: Optional[aiohttp.ClientSession] = None,
     ):
-        super().__init__(synthesizer_config)
+        super().__init__(synthesizer_config, aiohttp_session)
         # Instantiates a client
         azure_speech_key = azure_speech_key or getenv("AZURE_SPEECH_KEY")
         azure_speech_region = azure_speech_region or getenv("AZURE_SPEECH_REGION")

diff --git a/vocode/streaming/synthesizer/coqui_tts_synthesizer.py b/vocode/streaming/synthesizer/coqui_tts_synthesizer.py
@@ -2,6 +2,7 @@
 from concurrent.futures import ThreadPoolExecutor
 import logging
 from typing import Optional
+import aiohttp
 from pydub import AudioSegment
 import numpy as np
 import io
@@ -28,8 +29,9 @@ def __init__(
         self,
         synthesizer_config: CoquiTTSSynthesizerConfig,
         logger: Optional[logging.Logger] = None,
+        aiohttp_session: Optional[aiohttp.ClientSession] = None,
     ):
-        super().__init__(synthesizer_config)
+        super().__init__(synthesizer_config, aiohttp_session)
 
         from TTS.api import TTS
 

diff --git a/vocode/streaming/synthesizer/eleven_labs_synthesizer.py b/vocode/streaming/synthesizer/eleven_labs_synthesizer.py
@@ -29,8 +29,9 @@ def __init__(
         self,
         synthesizer_config: ElevenLabsSynthesizerConfig,
         logger: Optional[logging.Logger] = None,
+        aiohttp_session: Optional[aiohttp.ClientSession] = None,
     ):
-        super().__init__(synthesizer_config)
+        super().__init__(synthesizer_config, aiohttp_session)
 
         import elevenlabs
 

diff --git a/vocode/streaming/synthesizer/factory.py b/vocode/streaming/synthesizer/factory.py
@@ -1,6 +1,7 @@
 import logging
 from typing import Optional
 import typing
+import aiohttp
 
 from vocode.streaming.models.synthesizer import (
     AzureSynthesizerConfig,
@@ -31,22 +32,39 @@ def create_synthesizer(
         self,
         synthesizer_config: SynthesizerConfig,
         logger: Optional[logging.Logger] = None,
+        aiohttp_session: Optional[aiohttp.ClientSession] = None,
     ):
         if isinstance(synthesizer_config, GoogleSynthesizerConfig):
-            return GoogleSynthesizer(synthesizer_config, logger=logger)
+            return GoogleSynthesizer(
+                synthesizer_config, logger=logger, aiohttp_session=aiohttp_session
+            )
         elif isinstance(synthesizer_config, AzureSynthesizerConfig):
-            return AzureSynthesizer(synthesizer_config, logger=logger)
+            return AzureSynthesizer(
+                synthesizer_config, logger=logger, aiohttp_session=aiohttp_session
+            )
         elif isinstance(synthesizer_config, ElevenLabsSynthesizerConfig):
-            return ElevenLabsSynthesizer(synthesizer_config, logger=logger)
+            return ElevenLabsSynthesizer(
+                synthesizer_config, logger=logger, aiohttp_session=aiohttp_session
+            )
         elif isinstance(synthesizer_config, PlayHtSynthesizerConfig):
-            return PlayHtSynthesizer(synthesizer_config, logger=logger)
+            return PlayHtSynthesizer(
+                synthesizer_config, logger=logger, aiohttp_session=aiohttp_session
+            )
         elif isinstance(synthesizer_config, RimeSynthesizerConfig):
-            return RimeSynthesizer(synthesizer_config, logger=logger)
+            return RimeSynthesizer(
+                synthesizer_config, logger=logger, aiohttp_session=aiohttp_session
+            )
         elif isinstance(synthesizer_config, GTTSSynthesizerConfig):
-            return GTTSSynthesizer(synthesizer_config, logger=logger)
+            return GTTSSynthesizer(
+                synthesizer_config, logger=logger, aiohttp_session=aiohttp_session
+            )
         elif isinstance(synthesizer_config, StreamElementsSynthesizerConfig):
-            return StreamElementsSynthesizer(synthesizer_config, logger=logger)
+            return StreamElementsSynthesizer(
+                synthesizer_config, logger=logger, aiohttp_session=aiohttp_session
+            )
         elif isinstance(synthesizer_config, CoquiTTSSynthesizerConfig):
-            return CoquiTTSSynthesizer(synthesizer_config, logger=logger)
+            return CoquiTTSSynthesizer(
+                synthesizer_config, logger=logger, aiohttp_session=aiohttp_session
+            )
         else:
             raise Exception("Invalid synthesizer config")
diff --git a/vocode/streaming/synthesizer/google_synthesizer.py b/vocode/streaming/synthesizer/google_synthesizer.py
@@ -5,6 +5,7 @@
 import os
 import wave
 from typing import Any, Optional
+import aiohttp
 
 from vocode import getenv
 
@@ -24,13 +25,13 @@
 
 
 class GoogleSynthesizer(BaseSynthesizer[GoogleSynthesizerConfig]):
-
     def __init__(
         self,
         synthesizer_config: GoogleSynthesizerConfig,
         logger: Optional[logging.Logger] = None,
+        aiohttp_session: Optional[aiohttp.ClientSession] = None,
     ):
-        super().__init__(synthesizer_config)
+        super().__init__(synthesizer_config, aiohttp_session)
 
         from google.cloud import texttospeech_v1beta1 as tts
         import google.auth

diff --git a/vocode/streaming/synthesizer/gtts_synthesizer.py b/vocode/streaming/synthesizer/gtts_synthesizer.py
@@ -1,6 +1,7 @@
 import asyncio
 from concurrent.futures import ThreadPoolExecutor
 import logging
+import aiohttp
 from pydub import AudioSegment
 from typing import Optional
 from io import BytesIO
@@ -21,8 +22,9 @@ def __init__(
         self,
         synthesizer_config: GTTSSynthesizerConfig,
         logger: Optional[logging.Logger] = None,
+        aiohttp_session: Optional[aiohttp.ClientSession] = None,
     ):
-        super().__init__(synthesizer_config)
+        super().__init__(synthesizer_config, aiohttp_session)
 
         from gtts import gTTS
 

diff --git a/vocode/streaming/synthesizer/play_ht_synthesizer.py b/vocode/streaming/synthesizer/play_ht_synthesizer.py
@@ -27,8 +27,9 @@ def __init__(
         api_key: Optional[str] = None,
         user_id: Optional[str] = None,
         logger: Optional[logging.Logger] = None,
+        aiohttp_session: Optional[aiohttp.ClientSession] = None,
     ):
-        super().__init__(synthesizer_config)
+        super().__init__(synthesizer_config, aiohttp_session)
         self.synthesizer_config = synthesizer_config
         self.api_key = api_key or getenv("PLAY_HT_API_KEY")
         self.user_id = user_id or getenv("PLAY_HT_USER_ID")

diff --git a/vocode/streaming/synthesizer/rime_synthesizer.py b/vocode/streaming/synthesizer/rime_synthesizer.py
@@ -31,8 +31,9 @@ def __init__(
         self,
         synthesizer_config: RimeSynthesizerConfig,
         logger: Optional[logging.Logger] = None,
+        aiohttp_session: Optional[aiohttp.ClientSession] = None,
     ):
-        super().__init__(synthesizer_config)
+        super().__init__(synthesizer_config, aiohttp_session)
         self.api_key = getenv("RIME_API_KEY")
         self.speaker = synthesizer_config.speaker
         self.sampling_rate = synthesizer_config.sampling_rate

diff --git a/vocode/streaming/synthesizer/stream_elements_synthesizer.py b/vocode/streaming/synthesizer/stream_elements_synthesizer.py
@@ -27,8 +27,9 @@ def __init__(
         self,
         synthesizer_config: StreamElementsSynthesizerConfig,
         logger: Optional[logging.Logger] = None,
+        aiohttp_session: Optional[aiohttp.ClientSession] = None,
     ):
-        super().__init__(synthesizer_config)
+        super().__init__(synthesizer_config, aiohttp_session)
         self.voice = synthesizer_config.voice
 
     async def create_speech(